Create CloudWatch Alarm

Auto Scaling uses CloudWatch alarms to trigger scaling actions. An alarm watches a single metric (for example: CPU utilization) over a time period you set, and performs one or more actions based on the value of the metric relative to a given threshold. CloudWatch alarms will not invoke actions just because they are in a particular state. For more information, see the Eucalyptus CloudWatch documentation in the Eucalyptus User Guide.

To create a CloudWatch alarm

  1. Enter the name for your alarm in the Name text box.
  2. Type a description for your alarm in the Description text box.
  3. Select trigger dimensions, if not already pre-populated with the selected resource such as an image or scaling group. A dimension is a name-value pair that uniquely identifies a metric; for example: "Scaling group" = "myscalinggroup".
  4. When Create Alarm is launched from a resources' detail page, pre-defined metrics specific to those resources display. Select a metric from the drop-down list box. A metric is a time-ordered set of data points - for example, CPU utilization or volume write ops. You can get metric data from Eucalyptus cloud resources (like instances or volumes), or you can publish your own set of custom metric data points to CloudWatch. You then retrieve statistics about those data points as an ordered set of time-series data.
  5. Select a statistic from the Trigger alarm when the drop-down list box. A statistic is computed aggregation of metric data over a specified period of time; valid values are average, minimum, maximum, sum, and data samples.
  6. Select a comparison operator for the statistic value.
  7. Select a trigger threshold value. This is value is used in combination with the comparison operator and the measurement period and time length to determine whether the alarm should be triggered.
  8. Select the number of measurement periods and the period lengths using the controls in the for at least and consecutive period(s) of areas of the dialog box. Periods define the time period over which the metric value is compared to the trigger threshold, as well as how many consecutive periods the trigger threshold must be breached before the alarm is triggered.
    As you specify the values in this section, the chart that follows display the graphical representation of that statistic data.
  9. In the Actions section of the Create alarm window, specify the action to be taken when an alarm is triggered.
    Note: This section does not display if creating alarms is launched from the create scaling policy workflow.
    1. The Action type is pre-populated with the resource activity associated with the alarm.
    2. Select the state you want the alarm to take action from the When alarm state is drop-down list box.
    3. If launched from a scaling group, the name of that scaling group displays as the default selection in the For scaling group drop-down list box. Select a different scaling group, if needed.
      When a scaling group is selected, the Execute scaling policy drop-down list box displays with only the policies associated with the selected scaling group.
    4. Choose a policy to execute from the Execute scaling policy drop-down list box.
      Note: If there is already an action associated with this alarm, the policy selection is not applicable.
      Note: If all scaling policies are already associated with actions, or there are none defined, the policy selection displays, "No policies available."
  10. To save your new alarm, click the Create Alarm button.
For more details about CloudWatch alarm properties, see AWS Resource Types Reference for Alarms.